Specifications
NPU
Mobilint ARIES — 8 NPU cores at 1.25 GHz
AI Performance
80 TOPS (INT8)
Memory Type
LPDDR4X, onboard
Memory Capacity
16 GB / 32 GB
Memory Bandwidth
66.7 GB/s
Flash Memory
32 MB
Host Interface
PCI Express Gen4, 8-lane (MXM slot)
Form Factor
MXM module, Type A: 82 (L) x 70 (W) mm
Weight
110 g
Thermal Design Power
25 W
Input Voltage
DC +12 V
External Power Connector
Not required
Cooling System
Heatsink or fan cooling
Operating Temperature
-35 C to +65 C standard; -35 C to +75 C extended
UART
1 port
LED Indicators
2 (Power, Heartbeat)
Supported OS
Linux, Windows
Supported Frameworks
PyTorch, TensorFlow, TFLite, ONNX, Keras
SDK
Mobilint SDK qb
Price
Quote upon request
Availability
Available
Overview
The MLA100 MXM takes the ARIES NPU and packages it as an 82 x 70 mm MXM module that drops into any MXM-compatible carrier board. It delivers 80 TOPS of INT8 inference from eight NPU cores at 1.25 GHz with 16 GB or 32 GB of onboard LPDDR4X at 66.7 GB/s, all inside a 25 W envelope — and with no external power connector, since the module is fed from the host slot.
The module is specified for extended temperature operation: -35 C to +65 C as standard, with a -35 C to +75 C extended range for vehicle-mount and outdoor deployments. That temperature envelope, combined with the 82 x 70 mm footprint and 25 W power draw, targets industrial embedded systems where space, airflow and power are all constrained.
Typical integration targets published by Mobilint include industrial controllers, inspection systems and factory-automation equipment; UAV and drone payloads running vision and language workloads; AIoT devices and home appliances needing on-device voice, vision and language processing without cloud connectivity; and interactive kiosks requiring real-time product and user recognition.
Key Benefits
80 TOPS in a standard MXM slot — no carrier redesign required. 25 W with no auxiliary power connector. Extended -35 to +75 C range for industrial and vehicle-mount use. Same ARIES software stack as the MLA100/MLA400 PCIe cards.
Applications
Industrial controllers and factory automation; machine-vision inspection systems; UAV and robotics payloads; AIoT appliances and smart home robots; interactive kiosks; medical and analytical instruments requiring on-device inference.
Request a Quote — MOBILINT ARIES MLA100 MXM — 80 TOPS EDGE AI MODULE FOR EMBEDDED SYSTEMS
QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.
Get Your Quote →